Vitrectomy for lamellar holes with epiretinal proliferation: a gain worth having, from two separate pools

Design

Systematic review and meta-analysis, PROSPERO registered, arms pooled separately

Population

610 eyes in 17 studies with lamellar macular holes and epiretinal proliferation (328 operated, 282 observed)

Primary outcome

Change in best-corrected visual acuity in logMAR

Effect

Vitrectomy -0.170 logMAR (95% CI -0.254 to -0.086), I-squared 0 per cent; observation -0.029 (95% CI -0.118 to 0.176)

Seventeen studies and 610 eyes with lamellar macular holes associated with epiretinal proliferation were pooled, 328 eyes operated across 14 studies and 282 observed across five. Vitrectomy was associated with a mean improvement of 0.170 logMAR (95% CI -0.254 to -0.086, P < 0.001) with heterogeneity of zero. Observed eyes changed by -0.029 logMAR (95% CI -0.118 to 0.176, P = 0.695), which is to say not at all.

The design detail decides how much this is worth. Comparative data were too sparse to pool directly, so the two arms were meta-analysed separately and then set side by side. That is not a randomised comparison and it carries every selection difference between eyes chosen for surgery and eyes chosen for follow-up. The safety question came out reassuring but imprecise: progression to full-thickness macular hole was 7.5 per cent (95% CI 4.0 to 13.6) after vitrectomy and 4.2 per cent (1.4 to 11.8) after observation, and bias-adjusted estimates widened both to 12.4 per cent and 10.1 per cent with heavily overlapping intervals.

In clinic this supports offering surgery to a symptomatic patient with supportive optical coherence tomography features, rather than defaulting to watchful waiting on the grounds that lamellar holes are benign. It does not support operating on an asymptomatic eye to prevent progression, because the progression figures do not separate.

• Anchor the decision on symptoms and OCT morphology, not on the label lamellar hole alone.
• Quote a gain of roughly one to two Snellen lines on average, with wide individual variation.
• Do not promise that surgery lowers the risk of a full-thickness hole; the intervals overlap.
• Note that the zero heterogeneity applies to the surgical pool only, and reflects consistency, not certainty.
• Re-image before listing, since epiretinal proliferation can be mistaken for ordinary epiretinal membrane.

The statistics, in plain English

A negative logMAR change means better acuity, so -0.170 is a real improvement and its interval stays entirely on the beneficial side. The observational pool's interval runs from -0.118 to 0.176 and straddles zero, meaning no detectable change either way. Because the two arms were pooled separately rather than compared within studies, the difference between them is a comparison of averages from different populations and is weaker than the tidy numbers suggest.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.